Evaluation of prostatic cancer histology and grade distribution: experience with the Colorado Central Cancer Registry.

Although well-defined grading schemes for prostatic adenocarcinoma have been developed, they are not yet universally accepted by practicing surgical pathologists. The complexity of these schemes often leads surgical pathologists to develop their own modified Broders scheme. This study compared the grade distribution as obtained by a community of surgical pathologists with that obtained using the National Prostatic Cancer Project (NPCP) and Gleason grading schemes. In 1978, 308 cases of prostate cancer were reported to the Colorado Central Cancer Registry (CCCR) from the Denver Standard Metropolitan Statistical Area. Two hundred eighteen of these cases were regraded. The grade distribution as reported by the CCCR revealed a predominance of low-grade tumors (grade I-41%, grade II-28%, grade III-17%, grade IV-3%). Regrading of these same cases revealed a shift to higher grades (NPCP: grade I-14%, grade II-11%, grade III-37%, grade IV-29%; Gleason: pattern scores less than 6-30%, score 6-24%, score 7-12%, score 8-11%, score 9-10%, score 10-4%). Twenty-one cases of histologic variants which were not originally diagnosed were also noted (six mucinous, 12 ductal, two endometroioid, one squamous). There were 18 cases in which no evidence of carcinoma was confirmed. These results suggest that there is a tendency to underestimate grade and potentially malignant behavior when well-defined prostatic cancer grading schemes are not applied.